Optical pattern recognition based on image correlation is highly useful in automatic inspection, testing, surveillance and robotic vision. Image correlation involves the detailed comparison of an input image with a reference image. This detailed comparison can yield information regarding the similarity of the input image and the reference image. This technique may be useful for identification of radar or sonar images or for identification of parts via a machine vision system.
Coherent optical systems can provide high speed processing of images with large space-bandwidth products. It would be highly useful to develop optical correlator systems that produce correlation signals with large peak intensity, narrow correlation bandwidth, small auto-correlation sidelobes and good discrimination sensitivity.